Constrained neural adaptive PID control for robot manipulators

The problem of designing an analytical gain tuning and stable PID controller for nonlinear robotic systems is a long-lasting open challenge. This problem becomes even more intricate if unknown system dynamics and external disturbances are involved. This paper presents a novel adaptive neural-based control design for a robot with incomplete dynamical modeling and facing disturbances based on a simple structured PID-like control. Radial basis function neural networks are used to estimate uncertainties and to determine PID gains through a direct Lyapunov method. The controller is further augmented to provide constrained behavior during system operation, while stability is guaranteed by using a barrier Lyapunov function. The paper provides proof that all signals in the closed-loop system are bounded while the constraints are not violated. Finally, numerical simulations provide a validation of the effectiveness of the reported theoretical developments. (C) 2020 The Franklin Institute. Published by Elsevier Ltd. All rights reserved.